Blog

AI Recommendations

How checkout recommendations increase average order value by showing relevant suggestions at the right moment, with pricing tied only to additional sales.

Published April 10, 2026

Average order value often grows through small decisions made near the end of checkout. A customer adds the main item, reaches the final step, and then sees one more relevant product, add-on, or upgrade that fits the purchase. That moment matters because the buyer already has intent. The question is whether the business presents the next best option clearly enough.

pi-square supports recommendations during checkout to help businesses increase order value without making the journey feel forced or cluttered. The goal is not to distract the customer. It is to show useful suggestions at the point where they are most likely to be accepted.

Why checkout recommendations matter

Many businesses lose easy revenue because they stop selling once the customer adds the first item. In practice, customers are often open to buying more if the suggestion is relevant and timely.

That can include:

• add-ons that complement the main item
• higher-value alternatives
• related products that fit the same use case
• small extras that improve the overall order

When shown properly, these suggestions increase average order value without requiring more traffic.

Recommendations work best near purchase

There is a big difference between showing generic product suggestions early in browsing and showing targeted recommendations during checkout. At checkout, the customer has already made a decision to buy. That makes the recommendation slot more commercially important.

For a seller, this can improve:

• average order value
• revenue per customer
• attachment rate for add-ons
• visibility for complementary products
• overall efficiency of existing traffic

This is especially useful for businesses that already have decent conversion and want to grow revenue without depending only on acquiring more customers.

What this looks like in practice

Different business types can use recommendations in practical ways:

• a bakery can suggest candles, greeting cards, or celebration add-ons with a cake order
• a grocery seller can suggest staples or fast-moving products commonly bought together
• an event business can suggest upgrades, add-on services, or premium passes
• a service seller can suggest complementary services or follow-up packages

The value comes from relevance. Recommendations should feel connected to the order, not random.

Increase average order value without adding customer friction

The strongest recommendation systems do not rely on aggressive upselling. They work because they reduce the effort required for the customer to discover something else worth adding.

That matters because average order value usually improves when:

• the additional item is easy to understand
• the suggestion appears at the right time
• the customer does not need to restart the buying process
• the added value is obvious

Checkout is one of the best places to do this because attention is already focused on completing the purchase.

A pricing model tied to additional sales

One useful part of the pi-square recommendations feature is the commercial model. The seller pays only 10% of additional sales generated through recommendations.

That structure is important because it aligns cost with performance:

• there is no need to pay a large fixed fee just to try the feature
• the cost is tied to measurable upside
• the seller shares a small portion only when recommendations actually generate more revenue

For many small businesses, that makes the feature easier to adopt because it reduces upfront risk.

Why this is useful for small businesses

Small businesses often do not have large marketing budgets or dedicated merchandising teams. They still need ways to improve revenue from existing customer demand. Checkout recommendations help by improving monetisation of orders already in progress.

Instead of depending only on more traffic, the business can improve the value of each successful order.

That is often a smarter next step when:

• traffic is limited
• margins matter
• repeat buyers already trust the business
• product catalog depth allows meaningful add-ons

Use recommendations where they fit naturally

Not every catalog needs the same recommendation strategy. The feature is most useful when the business has complementary products, optional upgrades, or clear cross-sell opportunities.

The right approach is usually:

1. identify which products are commonly bought together
2. place those suggestions near checkout
3. monitor which recommendations actually lift order value
4. refine over time based on performance

That turns recommendations into a practical revenue lever instead of a decorative feature.

Final thought

Recommendations during checkout help businesses increase average order value at one of the most commercially important points in the customer journey. When the suggestions are relevant, the experience stays helpful for the buyer and more profitable for the seller.

pi-square makes that easier to adopt by tying the fee to results. If recommendations generate additional sales, the seller pays only 10% of that uplift. If they do not, there is no extra performance charge. That makes the feature easier to test, evaluate, and scale.

FAQ

What are checkout recommendations?

They are product or add-on suggestions shown during checkout to encourage relevant extra purchases before the order is completed.

How do recommendations increase average order value?

They help customers discover useful add-ons, upgrades, or related products at the point of purchase, which increases revenue per order.

When does the seller pay for the recommendations feature?

The seller pays only 10% of additional sales generated through recommendations.

Is this useful only for large ecommerce businesses?

No. Small businesses can benefit significantly because improving order value from existing traffic is often more efficient than chasing more traffic first.